Plan and apply separation in pipelines in Terraform - Step-by-Step Execution

Process Flow - Plan and apply separation in pipelines

Start Pipeline

↓

Terraform Plan Stage

↓

Review Plan Output

↓

Approve

↓

Terraform Apply

↓

Deploy Infrastructure

↓

End Pipeline

The pipeline first runs a plan to show changes, waits for approval, then applies changes to deploy infrastructure.

Execution Sample

Terraform

terraform plan -out=tfplan
# Review plan output
terraform apply tfplan

This code runs a plan to preview changes and then applies them if approved.

Process Table

Step	Action	Input/Condition	Result	Next Step
1	Run terraform plan	No prior plan	Plan file 'tfplan' created showing changes	Review plan output
2	Review plan output	Plan file exists	Changes understood, decision needed	Approve or Reject
3	Decision: Approve?	User input	If yes, proceed; if no, stop	Apply or Stop
4	Run terraform apply	Approved plan file	Infrastructure updated as planned	Deploy infrastructure
5	Deploy infrastructure	Apply successful	Infrastructure live and configured	End pipeline
6	Stop pipeline	Rejected plan	No changes applied	End pipeline

💡 Pipeline ends after apply or stop based on approval decision

Status Tracker

Variable	Start	After Plan	After Approval	After Apply	Final
plan_file	none	tfplan created	tfplan approved or rejected	tfplan applied or discarded	none or applied state
pipeline_status	started	planning	waiting approval	applying or stopped	completed or stopped

Key Moments - 3 Insights

Why do we separate 'plan' and 'apply' stages in the pipeline?

Separating plan and apply lets us review changes before making them live, as shown in execution_table rows 2 and 3.

What happens if the plan is rejected during approval?

The pipeline stops without applying changes, as shown in execution_table row 6.

Why do we save the plan output to a file before applying?

Saving the plan ensures the exact reviewed changes are applied later, preventing unexpected updates (execution_table row 1 and 4).

Visual Quiz - 3 Questions

Look at the execution table, what is the result of step 1?

A Infrastructure updated as planned

B Plan file 'tfplan' created showing changes

C No changes applied

D User input required

💡 Hint

Check the 'Result' column for step 1 in the execution_table

At which step does the pipeline wait for user approval?

A Step 3

B Step 2

C Step 4

D Step 6

💡 Hint

Look at the 'Action' and 'Next Step' columns around step 3 in the execution_table

If the plan is rejected, what is the pipeline status after step 6?

A applying

B completed

C stopped

D planning

💡 Hint

Refer to 'pipeline_status' in variable_tracker after step 6

Concept Snapshot

Terraform pipelines separate 'plan' and 'apply' stages.
Plan stage previews changes and saves to a file.
Approval is required before applying.
Apply stage uses the saved plan to update infrastructure.
This prevents unexpected changes and improves safety.

Full Transcript

In Terraform pipelines, we first run 'terraform plan' to see what changes will happen. This plan is saved to a file. Then, the pipeline waits for approval to make sure the changes are correct. If approved, 'terraform apply' uses the saved plan to update the infrastructure exactly as planned. If rejected, the pipeline stops and no changes are made. This separation helps avoid mistakes and keeps infrastructure safe.

Practice

1. What is the main benefit of separating terraform plan and terraform apply in pipelines?

easy

A. It speeds up the deployment by combining steps

B. It skips the need for state files

C. It automatically fixes errors during apply

D. It allows you to review changes before applying them

Solution

1. Step 1: Understand the purpose of terraform plan

   This command shows what changes Terraform will make without applying them.

2. Step 2: Understand the purpose of terraform apply

   This command applies the changes to the infrastructure based on the plan.

3. Final Answer:

   It allows you to review changes before applying them -> Option D

4. Quick Check:

   Plan shows changes first = B [OK]

Hint: Plan shows changes, apply makes them live [OK]

Common Mistakes:

• Thinking plan applies changes
• Believing apply skips state files
• Assuming steps run faster combined

2. Which command correctly saves a Terraform plan to a file named myplan.tfplan?

easy

A. terraform apply -out=myplan.tfplan

B. terraform save myplan.tfplan

C. terraform plan -out=myplan.tfplan

D. terraform plan save myplan.tfplan

Solution

1. Step 1: Identify the correct syntax for saving a plan

   The correct command uses terraform plan -out=filename to save the plan.

2. Step 2: Check each option

   terraform plan -out=myplan.tfplan matches the correct syntax exactly.

3. Final Answer:

   terraform plan -out=myplan.tfplan -> Option C

4. Quick Check:

   Plan with -out saves file = C [OK]

Hint: Use 'plan -out=' to save plans, not apply [OK]

Common Mistakes:

• Using apply instead of plan to save
• Using incorrect command like 'save'
• Missing the '-out=' flag

3. Given these commands run in order:
terraform plan -out=planfile
terraform apply planfile
What happens when you run terraform apply planfile?

medium

A. Terraform applies exactly the changes saved in planfile

B. Terraform ignores planfile and creates a new plan

C. Terraform applies changes but may differ from planfile

D. Terraform throws an error because planfile is not a valid command

Solution

1. Step 1: Understand the role of the saved plan file

   The plan file contains the exact changes Terraform will apply.

2. Step 2: Understand how terraform apply planfile works

   This command applies the saved plan exactly, without recalculating changes.

3. Final Answer:

   Terraform applies exactly the changes saved in planfile -> Option A

4. Quick Check:

   Apply saved plan = exact changes [OK]

Hint: Apply with plan file applies exact saved changes [OK]

Common Mistakes:

• Thinking apply recalculates plan
• Assuming apply ignores plan file
• Believing apply with plan file causes errors

4. You run terraform apply myplan.tfplan but get an error saying the plan file is invalid. What is the most likely cause?

medium

A. You forgot to run terraform init before apply

B. The plan file was created with a different Terraform version

C. You used terraform plan without -out flag

D. The plan file is empty because no changes were detected

Solution

1. Step 1: Identify compatibility issues with plan files

   Plan files are version-specific and may be invalid if Terraform versions differ.

2. Step 2: Check other options

   Running init is required but usually causes different errors; missing -out means no plan file saved; empty plan files do not cause invalid file errors.

3. Final Answer:

   The plan file was created with a different Terraform version -> Option B

4. Quick Check:

   Version mismatch causes invalid plan file error [OK]

Hint: Check Terraform versions match when using saved plans [OK]

Common Mistakes:

• Assuming missing init causes invalid plan file error
• Thinking empty plan files cause invalid file errors
• Confusing missing -out with invalid plan file

5. You want to implement a pipeline that separates planning and applying Terraform changes. Which sequence of commands ensures you can review the plan before applying exactly those changes?

hard

A. terraform plan -out=planfile && terraform apply planfile

B. terraform apply && terraform plan -out=planfile

C. terraform plan && terraform apply

D. terraform apply -out=planfile && terraform apply planfile

Solution

1. Step 1: Identify correct order for plan and apply separation

   First, run terraform plan -out=planfile to save the plan and review it.

2. Step 2: Apply the saved plan exactly

   Then, run terraform apply planfile to apply the reviewed changes exactly.

3. Final Answer:

   terraform plan -out=planfile && terraform apply planfile -> Option A

4. Quick Check:

   Plan with -out then apply saved plan = A [OK]

Hint: Plan with -out then apply saved plan file [OK]

Common Mistakes:

• Applying before planning
• Not saving plan with -out
• Using apply -out which is invalid